Advancements in the field of content distribution have led to development of various methods of delivering content using the Internet Protocol (IP). The content may include a live or on demand audio stream, a video stream, a multimedia content, and the like. Such content may be delivered from a Multimedia Video Programming Distributor (MPVD) using content distribution networks (CDNs).
In certain scenarios, when a lot of users attempt to access a content, such as a premium content or a new release of a Television series, there may be a lot of concurrent requests for such content from geographically diverse locations. Further, such requests may be received by the MPVD within a short amount of time. In such scenarios, the network of CDNs replicates the caches based on the increased number of user requests to access the content. However, such a replication of caches may result in degraded network performance and quality of service (QoS) parameters, such as delayed video starts, video start failures, and increased buffering time, and the like. In certain other scenarios, the content may be pushed to the last mile but such an operation may not be stochastic in nature and may not be customized to the users' requirements. This may lead also lead to reduced network performance and increased inefficiencies due to limited network resources. Furthermore, such a reduced performance may motivate the users to not access the content that may further lead to monetary losses for the content distributors.
It is therefore desirable to provide a mechanism for efficiently determining the demand of a particular content with respect to the requirements of the users and to ensure high QoS while delivering the content to the users.